Parallel high-dimensional multi-objective feature selection for EEG classification with dynamic workload balancing on CPU-GPU architectures.
Juan José EscobarJulio OrtegaJesús GonzálezMiguel DamasAntonio F. DíazPublished in: Clust. Comput. (2017)
Keyphrases
- feature selection
- multi objective
- high dimensional
- high dimensionality
- feature space
- graphics processing units
- classification accuracy
- parallel implementation
- classification models
- machine learning
- heterogeneous computing
- text classification
- support vector
- support vector machine
- dimensionality reduction
- feature vectors
- small sample
- parallel computing
- feature selection algorithms
- feature extraction
- parallel processing
- dimension reduction
- real time
- feature set
- evolutionary algorithm
- high dimension
- general purpose
- irrelevant features
- microarray data
- graphics processors
- parallel programming
- high dimensional data
- model selection
- multi objective optimization
- method for feature selection
- classification performances
- gene expression data
- data transfer
- eeg signals
- text categorization
- decision trees
- neural network
- discriminative features
- nearest neighbor
- parallel architectures
- eeg data
- multi core processors
- motor imagery
- multi class
- data points
- compute intensive